Fingerprint UUID is different on initial page load (and when refreshing without cache) #307
Comments
|
I found a Fiddle that was created a while back for another test with discrepancies. I used it to show the response on first load vs. subsequent loads. (I thought maybe it was the page I was integrating it in, but the Fiddle gives similar responses). At a precursory glance, adblock=false for the first load; however adblock=true for each subsequent load. I believe I saw a similar result in a previous issue. There is also a merge for fixing AdBlock detection (#147) where it was briefly discussed but I don't see a resolution (unless it's due to a delay in AdBlock element creation). FYI I am using Adblock Plus. Also to note, the fingerprint UUID I'm given on the Fiddle is not the same returned via my code integration. Not sure if the UUID building has changed between these times. Below are the Fiddle output results: FIRST LOAD SUBSEQUENT LOADS: |
|
Could you please try to delay fingerprinting for a few seconds/milliseconds. Does that make fingerprinting stable? |
|
@jonashaag Yes indeed; adding a delay of ~50ms seems to consistently net the same fingerprint now. |
|
Ok, closing then. This is nothing we can fix. |
|
Good news. Perhaps an entry in the FAQ should be usefull ? |
|
There's a hint here in the README already but I admit it's hard to find: https://github.com/Valve/fingerprintjs2#userdefinedfonts-option I'll add a note to the code sample. If you have other suggestions with respect to documentation, feel free to send a PR! |
|
I am facing the same problem even after added few second of delay. |
|
Please post your fingerprints using this jsfiddle https://jsfiddle.net/efer6pgy/ |
|
attached files are the two fingerprints and it's detailed components, one is generated by normal (non-incognito) mode and another is in the incognito mode of Mozilla Firefox browser : fingerpint : 0df45544f41c415bf75e91ca2d24afae fingerpint : 643f7718f6abb6a195f24bfb200224de |
|
What are the two configurations exactly? "Normal windows"? Are you comparing output of a non-incognito and incognito mode browser? Are you using the master version of fingerprintjs? |
|
Yes, one is incognito and another is non-incognito mode browser. Also, above fingerprint were generated by above link you provided me. Thanks in advance. |
|
OK, please try this https://jsfiddle.net/L2gLq4rg/ |
|
It seems like the problem have been solved. But, I don't know the changes you have made, could you please explain me in details? Thank you! |
|
Hello @jonashaag, |
|
I used the master version of fingerprintjs |
|
Thank you @jonashaag 👍 |
|
Would it make sense to add a time-to-load to each source in readme? If latency is ever a requirement, it sort of sucks to learn about slow sources by trial and error. |
|
PR welcome |
GrosSacASac
pushed a commit
to GrosSacASac/fingerprintjs2
that referenced
this issue
Sep 19, 2018
jonashaag
pushed a commit
that referenced
this issue
Sep 20, 2018
* use valid html * rename oncompleteTimeout * refactor: remove unnecessary keys reassignement * Give example of setTimeout fixes #307 #254 * Refactor, avoid innerHTML, fingerprint on load, avoid escapeHTML hack * 1.8.0 -> 1.8.1 * 1.8.0 -> 1.8.2 * 1.8.0 -> 1.8.2 * revert image removal * drop node 5,7,9 * avoid fingerprint, in case we click before the idleCallback * recommen requestIdleCallback in the example * add a comment to go read the note in the README.
|
Even after adding a delay of 2000ms, its showing different fingerprints on initial load and when refreshing. Also, different fingerprints are generated in normal mode and in incognito mode (Browser: Microsoft Edge) . |
|
For me, I was importing the Impact font twice in my project. When i delete the extra font the hash became consistent. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
(I couldn't find any previous issues regarding this) This is happening with all major browsers on both Windows 10 and macOS High Sierra.
Upon the first page load, a particular fingerprint UUID is given. However, after every subsequent load, the fingerprint UUID is different (and will remain constant with each normal refresh).
However - if you shift-reload the page, the UUID goes back to the original one that was first given on first load.
Does this have something to do with items being cached? I can't get it to output a consistent UUID between first load and subsequent loads. (and obviously, the first load will generally be pretty important).
The text was updated successfully, but these errors were encountered: